Microvascular invasion (MVI) is a critical prognostic factor that significantly impacts postoperative outcomes in hepatocellular carcinoma (HCC). As the current gold standard for the diagnosis of MVI is based on the postoperative histopathological examination of whole slide images, accurate preoperative prediction of MVI status using magnetic resonance imaging (MRI) presents both a substantial clinical imperative and a significant challenge. In order to discover reliable MRI-based imaging biomarkers to support clinical decision making and enhance the interpretability of deep learning-based diagnostic models, we propose a novel interpretable MVI prediction framework in which the shared latent visual attributes are first learned and then used for potential imaging biomarker extraction and MVI diagnosis, respectively. To ensure that the visual attributes of these biomarkers are generalizable across diverse patients, the similarity constraints at the intra-patient level and the inter-patient level are enforced within the learned feature space, enabling intuitive biomarker discovery directly from the original image space. To guarantee semantic alignment between biomarkers and the characteristics of individual patients, we introduce a novel classification mechanism that directly links the alignment between each biomarker and patient-specific characteristics with the prediction, thereby ensuring a precise prediction of MVI. Furthermore, the interpretability of the model is enhanced by integrating a mask-based visual explanation method that highlights regions in patient images that correspond to the identified biomarkers. Extensive experiments on two MVI prediction datasets: HCC-WCH and HCC-ZSH unequivocally demonstrate our method's superior performance in both classification accuracy and interpretability.

Imbalanced small datasets are common scenarios in the field of machine learning for medical imaging, especially in real-world clinical applications. Many existing works focus on synthesize new images via data generation. However, generative methods cannot ensure reliability for medical images where categories cannot be easily distinguished, such as the pathologic complete response (pCR) evaluation via MRIs in cancer prognosis. Meanwhile, few-shot learning can deal with training with small datasets, but it depends on a balanced data distribution and a large number of image categories. In this paper, we propose an image similarity comparison classification network, referred to as Cross-Supervision Similarity Network (CSSN), using cross-supervision between class features and patch features. CSSN transforms the classification task into comparison task by calculating similarity scores at both patch and class scales, effectively training on imbalanced small datasets with limited categories. To balance the training difficulty of the two similarity branches, soft logarithmic supervision is used to construct soft labels between them. Through experiments on PCR-ISD, we observe significant performance improvements of 15% in F1 score, 6% in accuracy and 9 % in balanced accuracy over existing methods, indicating the superiority of our method in identifying minority classes and enhancing classification capabilities. Extensive experiments on three datasets and ablation experiments confirm the effectiveness and generalization ability of the proposed method. The source code is available at https://github.com/lxy-146/CSSN_TMI.

Microvascular invasion (MVI) is a critical determinant that substantially influences the postoperative prognosis of hepatocellular carcinoma (HCC). Accurate preoperative diagnosis of MVI using MRI imaging has far-reaching research implications. While deep learning models have demonstrated remarkable diagnostic performance, their intrinsic black-box nature poses significant challenges to further advancement. To address this limitation, we propose a novel, model-agnostic interpretation approach, the Biosignature Identification Framework (BIF), inspired by causal inference theory and the biological concept of biosignatures. Within BIF, the Biosignature Identification Module (BIM) operates in parallel with the prediction model, identifying key biosignatures and generating interpretations based on these biosignatures. Unlike conventional model-agnostic interpretation techniques, BIF uniquely offers definitive interpretations grounded in causal inference, thereby enhancing the accuracy and credibility of the interpretive process. Extensive experiments on a clinical dataset collected by Zhongshan Hospital demonstrate the interpretability and efficacy of BIF in preoperative MVI prediction for HCC.

Microvascular invasion (MVI) is an independent risk factor for postoperative recurrence of hepatocellular carcinoma (HCC). Preoperative MVI grading is beneficial for patients recovery and survival. However, preoperative MVI grading is primarily accomplished through magnetic resonance imaging (MRI), which is challenging due to the heterogeneity of tumors and the characteristics of MVI. In this paper, we propose a cascade network that extracts fine-grained information from multimodal MRI images to assist in accurate MVI grading. We extract fine-grained features from different modalities and integrate them using an attention-based module called Multimodal Fine-Grained generator (M-FG) to obtain finegrained features from multimodal MRIs. Extensive experiments show our MVI grading network achieved an accuracy of 0.77, up to 10% improvement compared to the comparative methods, which validates that our method effectively utilizes fine-grained features from different modalities and improves performance of MVI grading. The codes are available at https://github.com/lxy-146/FG_MVInet

This article analyzes a switched-capacitor-based seven-level inverter modulated by a hybrid of level- and phase-shifted pulsewidth modulation (PWM) algorithm in detail. The optimal design of circuit parameters is also presented to ensure this hybrid PWM algorithm gives full play to its advantages. The two capacitors employed in this inverter operate alternately in charging and discharging states at carriers' frequency of this PWM algorithm, to provide different output voltage levels. With reasonable parameter design as addressed in this article, both capacitors can be fully charged every time and the phenomenon of capacitors' voltage ripple accumulation existed in other works is no longer found in this seven-level inverter. All transistors employed in this inverter are switched under low switching stress, resulting in less switching loss. Both simulation and experimental results are provided to demonstrate the correctness of theoretical analysis, and the maximum measured efficiency is up to 96.4%.

When a parallel manipulator suffers from failures, its performance can be significantly affected. Thus, fault tolerance is essential for task-critical applications or applications in which maintenance is hard to implement. In this paper, we consider three types of common strut failures corresponding to stuck joints, unactuated actuators, or the complete loss of struts, respectively. The impacts of different failures on the kinematics of a manipulator are examined, and the task space redundancy and kinematic redundancies are used to help overcome these failures. In addition, local measures of fault tolerance and their properties are analyzed. These measures can be helpful in architecture design and path planning

This paper develops methods for generating classes of orthogonal Gough-Stewart platforms (OGSPs). First, a new, two-parameter class of six-strut OGSPs which leads to isotropic manipulators are found. Next, this class is extended to include redundant Gough-Stewart platforms (GSPs). For an even number of struts, the same algorithm used to generate the six-strut case can be employed. For an odd number of struts, similar essential concepts are used to derive seven-strut and nine-strut OGSPs. Maximization of fault tolerance is implemented for a nine-strut isotropic OGSP. By exploiting invariant properties of the inverse Jacobian, new methods for favorably altering the center of gravity, strut attachment surface, and strut spatial distribution are developed.

Optimal geometric design is of key importance to the performance of a manipulator. First, this paper extends the work in Y. Yi, et al., (2004) to generate a class of isotropic Gough-Stewart platforms (GSPs) with an odd number of struts. Then, it develops methods for finding a highly fault tolerant GSP from that class. Two optimization criteria are considered, isotropy and fault tolerance. To meet the mission critical needs imposed by laser weapons applications, nine-strut isotropic GSPs that retain kinematic stability despite the loss of any three struts are found. First, we develop methods for generating a five parameter class of isotropic nine-strut GSPs. Next, new measures of fault tolerance are introduced and used to optimize the free parameter space. The optimized design is much more fault tolerant than the GSP currently baselined for the airborne laser.

This paper develops methods for generating classes of orthogonal Gough-Stewart platforms (OGSPs). First, a new, two-parameter class of six-strut OGSPs is found. For an even number of struts, the same essential concepts used to generate the six-strut case can be employed to generate OGSPs with eight struts, ten struts, etc. Next, by exploiting invariant properties of the inverse Jacobian, new methods for favorably altering the center of gravity, strut attachment surface, and strut spatial distribution are developed. As an illustrative example, these methods are applied to the design of the attachment of vibration isolation and precision pointing system for the space based laser.

In many applications, the manipulations require only part of the degrees of freedom (DOFs) of the end-effector, or some DOFs are more important than the rest. We name these applications prioritized manipulations. The end-effector's DOFs are divided into those which are critical and must be controlled as precisely as possible, and those which have loose specifications, so their tracking performance can be traded off to achieve other needs. In this paper, for the class of general constrained rigid multibody systems (including passive joints and multiple closed kinematic loops), we derive a formulation for partitioning the task space into major and secondary task directions, and finding the velocity and static force mappings that precisely accomplish the major task and optimize some secondary goals such as reliability enhancement, obstacle and singularity avoidance, fault tolerance, or joint limit avoidance. The major task and secondary goals need to be specified in term of velocities/forces. In addition, a framework is developed to handle two kinds of common actuator failures, torque failure and position failure, by reconfiguring the differential kinematics and static force models. The techniques are tested on a 6-DOF parallel robot. Experimental results illustrate that the approach is practical and yields good performance.